2013 IEEE International Symposium on Information Theory 2013
DOI: 10.1109/isit.2013.6620417
|View full text |Cite
|
Sign up to set email alerts
|

Low separation rank covariance estimation using Kronecker product expansions

Abstract: Abstract-This paper presents a new method for estimating high dimensional covariance matrices. Our method, permuted rank-penalized least-squares (PRLS), is based on Kronecker product series expansions of the true covariance matrix. Assuming an i.i.d. Gaussian random sample, we establish high dimensional rates of convergence to the true covariance as both the number of samples and the number of variables go to infinity. For covariance matrices of low separation rank, our results establish that PRLS has signific… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
5
0

Year Published

2014
2014
2021
2021

Publication Types

Select...
5
1
1

Relationship

0
7

Authors

Journals

citations
Cited by 10 publications
(5 citation statements)
references
References 33 publications
0
5
0
Order By: Relevance
“…Given our focus on the experiences of postdoctoral women specifically, we isolated our analysis to the 146 women in the sample who were currently in a postdoctoral position ( M age = 33.40, SD = 4.45). According to G ∗ Power, this sample size provided us with 92% power to detect medium effect sizes (two-tailed; Faul et al, 2007; Erdfelder et al, 2009). Of this sample, roughly a third (37%) had at least one child.…”
Section: Methodsmentioning
confidence: 99%
“…Given our focus on the experiences of postdoctoral women specifically, we isolated our analysis to the 146 women in the sample who were currently in a postdoctoral position ( M age = 33.40, SD = 4.45). According to G ∗ Power, this sample size provided us with 92% power to detect medium effect sizes (two-tailed; Faul et al, 2007; Erdfelder et al, 2009). Of this sample, roughly a third (37%) had at least one child.…”
Section: Methodsmentioning
confidence: 99%
“…The minimum number of respondents required was determined by an a priori power analysis (G*Power: Faul, Erdfelder, Lang, & Buchner, 2007). A total minimum sample size of 126 was detected (80% chance) as significant at the 5% level with a small effect size for the between-respondent comparisons.…”
Section: Methodsmentioning
confidence: 99%
“…Remark. It is worth noting that the minimized function (16) does not coincide with the EXIP approach, since the weighted matrix Q is not the inverse of the (asymptotic) covariance matrix of vec R − R e given by (13). An EXIP-based procedure could be applicable, but the equivalence between (14) and ( 16) would be lost and would require an alternating optimization scheme to solve the corresponding weighted covariance fitting.…”
Section: B Step 2: Covariance Fittingmentioning
confidence: 99%
“…The problem of covariance estimation with KP structure and potential linearly constrained factors has been widely explored in the Gaussian framework [6]- [10], notably by using the Extended Invariance Principle (EXIP) [11]. Furthermore, other approaches, dealing with high dimensional data, have been proposed in [12], [13].…”
Section: Introductionmentioning
confidence: 99%